Disulfides and boryl sulfides serve as both initiators and precatalysts in radical reductions of halides by an N-heterocyclic carbene×Borane

Article abstract of DOI:10.1002/adsc.201300752

Diphenyl disulfide (PhSSPh), a typical boryl monosulfide (diiPr-Imd-BH ₂SPh) and a typical boryl bis-sulfide [diMe-Imd-BH(SPh)₂] all serve as both initiators and precatalysts in the reduction of alkyl and aryl halides by readily available 1,3-dimethylimidazol-2-ylidene×borane (diMe-Imd-BH₃). The reactions are suggested to occur by a polarity reverse catalysis mechanism where in situ generated thiophenol is the active catalyst. Copyright

Full text of DOI:10.1002/adsc.201300752

COMMUNICATIONS
DOI: 10.1002/adsc.201300752
Disulfides and Boryl Sulfides Serve as both Initiators and
Precatalysts in Radical Reductions of Halides by an
N-Heterocyclic Carbene·Borane
Xiangcheng Pan,a Jacques Lalevꢀe,b Emmanuel Lacꢁte,c,d and Dennis P. Currana,*
a
Department of Chemistry, University of Pittsburgh, Pittsburgh, PA 15260, USA
E-mail: curran@pitt.edu
b
Institut de Science des Matꢀriaux de Mulhouse IS2M – LRC CNRS 7228 – ENSCMu-UHA, 15 rue Jean Starcky, 68057
Mulhouse Cedex, France
ICSN CNRS, Avenue de la Terrasse, 91198 Gif-sur-Yvette Cedex, France
Universitꢀ de Lyon, Institut de chimie de Lyon, UMR 5265 CNRS – Universitꢀ Lyon I – ESCPE Lyon, 43 Bd du 11
c
d
novembre 1918, 69616 Villeurbanne, France
Received: September 6, 2013; Published online: November 27, 2013
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201300752.
Abstract: Diphenyl disulfide (PhSSPh), a typical
boryl monosulfide (diiPr-Imd-BH2SPh) and a typical
boryl bis-sulfide [diMe-Imd-BHACTHNUTRGNEUNG(SPh)2] all serve as
both initiators and precatalysts in the reduction of
alkyl and aryl halides by readily available 1,3-dime-
sion. The difficulties with poor propagation can be
offset either by changing the reaction conditions
(heating to higher temperatures, adding more initia-
tor) or by adding a catalyst.
We have recently shown that various thiols catalyze
the reductions of halides by NHC-boranes.[5] As
shown in Figure 1a, a halide RX (R=alkyl or aryl;
X=Br or I), an NHC borane (NHC-BH3, usually 1,3-
dimethylimidazol-2-ylidene borane) and 5% thiophe-
nol are heated or irradiated in the presence of 10–
20% of a suitable chemical initiator such as di-tert-
butyl hyponitrite (t-BuON=NO-t-Bu) or di-tert-butyl
peroxide (t-BuOO-t-Bu). The rates of reactions and
thylimidazol-2-ylidene·borane
(diMe-Imd-BH3).
The reactions are suggested to occur by a polarity
reverse catalysis mechanism where in situ generated
thiophenol is the active catalyst.
Keywords: N-heterocyclic carbene·boranes; polarity
reverse catalysis; radical reduction
À
yields of reduced products R H with thiol added are
significantly increased compared to control experi-
ments with no thiol added.
The use of N-heterocyclic carbene·boranes (NHC-bor-
anes) as reagents for radical reductions has attractive
features[1] compared to standard reagents like tributyl-
tin hydride.[2] NHC-boranes are stable solids that are
easy to make and handle, and the boron-derived
products are readily removed from crude reaction
products.[3] There are also important differences in re-
activity; specifically, tin hydrides are much better hy-
drogen donors[4] than NHC-boranes.[1b,c]
The lower H-donor capability of NHC-boranes is
a double-edged sword. On the one edge, it is helpful
when slow radical reactions (additions or cyclizations,
for example) are being conducted because the H-
transfer reaction of the NHC-borane does not com-
pete against the target reaction as efficiently as H-
transfer from a tin hydride. On the other edge, it is
harmful because slow steps in radical chains can lead
to difficulties in chain propagation. Reactions may
Figure 1. Summary of thiol-catalyzed halide reductions by
progress only with difficulty towards complete conver- NHC-boranes.
3522
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 3522 – 3526

Disulfides and Boryl Sulfides Serve as both Initiators and Precatalysts in Radical Reductions
Propagation steps in a suggested chain mechanism molytic substitution of diphenyl disulfide by an NHC-
for such reactions are shown in Figure 1b. In the ab- boryl radical, Figure 2b, step (5). The other step of
sence of the thiol, the halogen abstraction step (1) is the chain is the same hydrogen atom abstraction step
fast, but the direct hydrogen transfer step (2) is slow, (4) that was shown in Figure 1. Because thiophenol is
so chains do not propagate well.[1e] When added, thio- generated in these reactions, we hypothesized that di-
phenol serves as the primary hydrogen donor in step phenyl disulfide might serve as a precatalyst in place
(3)[6] because it is much more reactive towards radi- of the thiol in the PRC reactions of Figure 1. Like-
cals than NHC-borane 1. The phenylthiyl radical wise, we wondered whether the boryl mono- and bis-
(PhSC) generated in step (3) rapidly abstracts a hydro- sulfide products of the reactions in Figure 2 (for ex-
gen atom from the starting NHC-borane in step (4).[5] ample, 2 and 4) would also serve as precatalysts in
This step results in both catalyst turnover (the PhSH PRC reductions through either thermal or photolytic
À
product) and chain transfer (the NHC-BH2C product). cleavage of their B S bonds. If PhSSPh, 2 and 4 did
In short, added thiol catalyzes the reduction be- serve as precatalysts, then we speculated that they
cause two fast steps (3) and (4) supercede the slow might come with an added bonus that no chemical ini-
step (2). This process is typically called polarity rever- tiator would be needed. This is because the reactions
sal catalysis[7] because a mismatched polarity in step in Figure 2 occur smoothly with light only.
(2) (both reactants are nucleophilic) is reversed to
two matching polarities (one nucleophilic, one elec- typical boryl sulfides each serve as substitutes for
trophilic) in steps (3) and (4). both the thiol catalyst and the chemical initiator in
Here we show that diphenyl disulfide and the two
In a separate study, we learned that boryl sulfides typical PRC reactions of NHC-boranes. The upshot is
could be conveniently prepared from NHC-boranes that two convenient new procedures for halide reduc-
and disulfides under radical conditions.[8] For example, tions have been identified.
sunlamp irradiation of 1 equiv. of 1,3-dimethylimida-
Table 1 summarizes preliminary experiments with
zol-2-ylidene borane 1 with 2 equiv. of diphenyl disul- adamantyl iodide (AdI) and NHC-borane reductant
fide at 458C provided boryl bis-sulfide 2 in 80% yield 1 that probed whether both the thiophenol and the
(Figure 2a). Pure boryl monosulfides were difficult to chemical initiator in the current PRC procedure[5]
make from 1 (in the best cases, the samples of mono- (Figure 1a) could be replaced with either a disulfide
sulfides were contaminated with 10–15% bis-sulfide).
However, heating of the N-isopropyl analog 3 with
1 equiv. of diphenyl disulfide in lab light provided
clean monosulfide 4 in 66% yield.
These reactions are thought to occur by two-step
radical chains in which the key propagation step is ho-
Table 1. Reductions with diphenyl disulfide and two boryl
sulfides used as both initiators and catalyst precursors.[a]
Entry
Halide
Additive
Time
Yield[b]
1
2
I
I
Br
I
I
I
I
I
I
PhSSPh
none
PhSSPh
PhSSPh
2
4
2
4
2
2
2
4
1 h
3 h
3 h
3 h
1 h
1 h
24 h
24 h
1 h
5 h
7 h
7 h
84%
13%
81%
97%
90%
92%
23%
22%
98%
84%
27%
49%
3
4[c]
5
6
7[d]
8[d]
9[c]
10
11[e]
12[e]
Br
I
I
[a]
PhSSPh was used at 10%, boryl sulfides 2 or 4 at 15%.
Determined by H NMR spectroscopy.
The sunlamp was replaced by a black light, 15 W, 357 nm
max.
The sunlamp as not used, the mixture was heated at
808C in ambient light.
[b]
[c]
1
[d]
[e]
Figure 2. Synthesis of boryl mono- and bis-sulfides and sug-
gested propagation steps of a radical chain mechanism.
NHC-borane 1 was omitted and the additive 2 or 4 was
also used as the reductant (1.1 equiv.).
Adv. Synth. Catal. 2013, 355, 3522 – 3526
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3523

Xiangcheng Pan et al.
COMMUNICATIONS
Table 2. Preparative reductions under sunlamp conditions
with 1 equiv. of diMe-Imd-BH3 1 and 10% PhSSPh (condi-
tions A) or 15% diMe-Imd-BHACHTNUGTRENUNG(SPh)2 2 (conditions B).
or a boryl sulfide under sunlamp conditions. Light-in-
duced cleavage of a disulfide should give two thiyl
radicals that can react in step (3) with the NHC-
borane to generate the needed thiol catalyst in situ.
Indeed, sunlamp irradiation of adamantyl iodide,
NHC-borane 1 and 10 mol% diphenyl disulfide in
benzene-d6 produced adamantane in 84% yield after
1
only 1 h as assessed by H NMR spectroscopy. A con-
trol experiment with no additive gave low conversion
of AdI, producing only a 13% yield of adamantane
(entry 2).
Under the sunlamp conditions, adamantyl bromide
was reduced in comparable yield to the iodide (81%),
though the reaction time was a little longer (3 h,
entry 3). The reaction of AdI with diphenyl disulfide
was also conveniently conducted with black light as
recommended by Ryu,[9] providing adamantane in
97% yield after 3 h at room temperature (entry 4).
We observed small resonances for boron sulfides in
the 11B NMR spectra of some of the crude products
with diphenyl disulfide. Thus, we replaced the disul-
fide by the readily made boron sulfides 2 and 4,
which would again serve as both initiators (by cleav-
À
age of the B S bond) and sources of the thiol cata-
lyst.
Indeed, boron bis-sulfide 2 and monosulfide 4 both
proved competent at promoting rapid (1 h) and clean
reduction of adamantyl iodide (entries 5 and 6, 90 and
92%) when used at the 15 mol% level. Control ex-
periments under thermal conditions gave much lower
conversions; only 22–23% AdH was formed even
after extended reaction times (24 h, entries 7 and 8).
Clearly the sunlamp is important. Again, black light
was effective a promoting the reduction of AdI with
1 equiv. of 1 and 10% of 2 at room temperature
(entry 9, 98%).
To probe whether 2 and 4 could also serve as the
stoichiometric reductant, we conducted two experi-
ments with adamantyl iodide where we omitted the
NHC-borane 1 and increased the amount of boryl sul-
fides 2 and 4 to 1.1 equiv. Adamantane was formed,
but more slowly and in lower yield than with 1 equiv.
of 1 and 15% of the boryl sulfides. Over 7 h, 2 pro-
duced adamantane in 27% yield while 4 gave 49%
yield (entries 11 and 12). Apparently, the boryl sul-
fides 2 and 4 are good at initiating the chain and gen-
erating the thiol catalyst, but they are not as efficient
as 1 at propagating the chain by H-transfer [Figure 2,
step (5)]. Thus, the less expensive and lower molecu-
[a]
Sunlamp irradiation of a mixture of the precursor,
1 (1 equiv.) and either 10% PhSSPh (conditions A) or
15% boryl bis-sulfide 2 (conditions B) in benzene unless
otherwise indicated.
[b]
[c]
Isolated yields after automated flash chromatography
unless otherwise indicated.
Yield determined by 1H NMR spectroscopy.
Boryl bis-sulfide 2 is readily made in pure form so
lar weight NHC-BH3 complex 1 is the preferred stoi- it was used as an additive for several preparative re-
chiometric reductant. ductions with 1 under sunlamp irradiation (conditions
We next conducted some preparative reductions to B). Good yields were observed in the reductions of
assess isolated yields after automated flash chroma- cholesterol iodide and bromide 7a, b to deoxycholes-
tography, and selected results are summarized in terol 8 (98 and 82%, entries 2, 3).
Table 2. With 10% PhSSPh (conditions A), reduction
of acetone glucosyl iodide 5 provided the 3-deoxyglu- occur at all unless a catalyst is added.[5] Aryl iodide 9a
cose diacetonide 6 in 67% isolated yield (entry 1). was reduced effectively to 10 with both PhSSPh and
Reductions of aryl halides and 1 generally do not
3524
asc.wiley-vch.de
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 3522 – 3526

Disulfides and Boryl Sulfides Serve as both Initiators and Precatalysts in Radical Reductions
boryl bis-sulfide 2 (88% and 91%, entries 6 and 8). In can serve a dual role as both initiators (presumably
À
À
contrast, the aryl bromide 9b gave only partial con- by photocleavage of the S S or B S bond) and in situ
version under both conditions, and reactions were sources of the thiol that is needed to catalyze the re-
stopped when the yield of product 10 was 50% and actions.
40%, respectively (entries 7 and 9).
Typical radical cyclizations can also be conducted,
as indicated by the results in entries 10–12. Reduction Experimental Section
of iodide 11a with PhSSPh and bromide 11b with 4
smoothly provided 12 in 98 and 94% yields, respec- General procedure for Conditions A (PhSSPh) in
tively (entries 10 and 11). Finally, Ueno–Stork cycliza- Table 2
tion of 1,4-diene 13 provided two of the four possible
isomers 14 in 82% isolated yield (entry 12). The sub-
stituents on C-4 and C-5 are trans, while two anomers
Diphenyl disulfide (PhSSPh, 0.01 mmol, 10 mol%, 2.2 mg)
was added to a solution of diMe-Imd-BH3 1 (12 mg,
0.11 mmol) and the substrate (0.1 mmol) in benzene
at C-2 are present in about a 3/1 ratio. This stereo-
chemical outcome is similar to cyclization of sub-
strates like 11 conducted with other radical-generat-
ing reagents.[10]
(0.5 mL). The colorless solution was charged to a NMR tube
and irradiated with a 275 W sunlamp at 608C for 1–7 h. The
mixture was cooled to room temperature, then the solvent
was evaporated and the crude product was purified by flash
chromatography.
Most of the reactions in Table 2 were conducted in
benzene for consistency with the study in Table 1.
However, benzene is not an attractive solvent for
practical applications of preparative radical reactions.
To identify suitable solvents for such uses, we reduced
cholesterol bromide 7b with 1 by the two standard
procedures in toluene (procedure A) and benzotri-
fluoride (PhCF3, procedure B),[11] then determined
isolated yields of 8 by automated flash chromatogra-
phy as usual. Reaction times (all 7 h) and yields of 8
in the two new solvents were comparable to each
other (PhCH3, 81%; PhCF3, 85%, entries 4 and 5) and
to the reaction in benzene (82%, entry 2). As expect-
ed then, the selection of benzene as the solvent is not
a crucial reaction variable, and it can be replaced by
safer aromatic solvents and probably other types of
solvents as well.
General procedures for Conditions B [diMe-Imd-
BHACTHNUGTRENNG(U SPh)2] in Table 2
diMe-Imd-BHACHTNUTRGNEUNG(SPh)2 (2, 0.015 mmol, 15 mol%, 4.9 mg) was
added to a solution of diMe-Imd-BH3 1 (12 mg, 0.11 mmol)
and the substrate (0.1 mmol) in benzene (0.5 mL). The col-
orless solution was charged to a NMR tube and irradiated
with a 275 W sunlamp at 608C for 1–10 h. The mixture was
cooled to room temperature, then the solvent was evaporat-
ed and the crude product was purified by flash chromatogra-
phy.
Acknowledgements
This work was supported by the US National Science Foun-
dation. EL and JL thank the French Agence Nationale de
Recherche, UHA, ENSCMu, CPE Lyon, Universitꢀ de Lyon,
the Institut Universitaire de France (IUF), and CNRS for
support.
Overall, the additives PhSSPh and diMe-Imd-
BHACHTUNGTRENNUNG(SPh)2 2 give comparable yields to each other and
to similar reductions under the published procedure
with thiol catalysts and added chemical initiators.[5]
Alkyl iodides and bromides give roughly comparable
results, though the yields with bromides benefit from
somewhat longer reaction times. In contrast, aryl io-
dides perform better than aryl bromides because reac-
tions with the bromides tend to stall before full con-
version.
In summary, the reaction conditions for a new ho-
molytic substitution reaction of disulfides by boryl
radicals (Figure 2) are rather similar to the conditions
for the recently reported reduction of halides by po-
larity reverse catalysis[5] (Figure 1). This is because
the chain mechanisms share a common step [Figure 1
and Figure 2, step (4)]. Based on this sharing, we re-
placed the previously used combination of a thiol cat-
alyst and a chemical initiator in the polarity reverse
catalysis reductions with either a disulfide or a boryl
mono- or bis-sulfide. Both types of reactions gave re-
duced products with comparable rates and yields to
the published procedure. Evidently then, these species
References
[1] a) J. C. Walton, M. Makhlouf Brahmi, L. Fensterbank,
E. Lacꢁte, M. Malacria, Q. Chu, S.-H. Ueng, A. Solo-
vyev, D. P. Curran, J. Am. Chem. Soc. 2010, 132, 2350–
2358; b) A. Solovyev, S.-H. Ueng, J. Monot, L. Fenster-
bank, M. Malacria, E. Lacꢁte, D. P. Curran, Org. Lett.
2010, 12, 2998–3001; c) S.-H. Ueng, A. Solovyev, X.
Yuan, S. J. Geib, L. Fensterbank, E. Lacꢁte, M. Malac-
ria, M. Newcomb, J. C. Walton, D. P. Curran, J. Am.
Chem. Soc. 2009, 131, 11256–11262; d) S.-H. Ueng, M.
Makhlouf Brahmi, ꢃ. Derat, L. Fensterbank, E. Lacꢁte,
M. Malacria, D. P. Curran, J. Am. Chem. Soc. 2008,
130, 10082–10083; e) S.-H. Ueng, L. Fensterbank, E.
Lacꢁte, M. Malacria, D. P. Curran, Org. Biomol. Chem.
2011, 9, 3415–3420.
[2] a) P. A. Baguley, J. C. Walton, Angew. Chem. 1998, 110,
3272–3283; Angew. Chem. Int. Ed. 1998, 37, 3072–3082;
b) A. Studer, S. Amrein, Synthesis 2002, 835–849;
Adv. Synth. Catal. 2013, 355, 3522 – 3526
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3525

Xiangcheng Pan et al.
COMMUNICATIONS
c) J. C. Walton, A. Studer, Acc. Chem. Res. 2005, 38,
794–802; d) H. Jasch, M. R. Hienrich, in: Encyclopedia
of Radicals in Chemistry, Biology and Materials, Vol. 2,
(Eds.: C. Chatgilialoglu, A. Studer), Wiley, New York,
NY, 2012, pp 529–560.
[7] a) D. Crich, X. S. Mo, J. Am. Chem. Soc. 1998, 120,
8298–8304; b) B. P. Roberts, Chem. Soc. Rev. 1999, 28,
25–35; c) C. Chatgilialoglu, Helv Chim Acta 2006, 89,
2387–2398.
[8] X. Pan, A.-L. Vallet, S. Schweizer, K. Dahbi, B. Del-
pech, N. Blanchard, B. Graff, S. J. Geib, D. P. Curran, J.
Lalevꢀe, E. Lacꢁte, J. Am. Chem. Soc. 2013, 135,
10484–10491.
[9] T. Kawamoto, T. Fukuyama, I. Ryu, J. Am. Chem. Soc.
2012, 134, 875–877.
[10] X. J. Salom-Roig, F. Denes, P. Renaud, Synthesis 2004,
1903–1928.
[11] a) J. J. Maul, P. J. Ostrowski, G. A. Ublacker, B. Lin-
clau, D. P. Curran, Modern Solvents in Organic Synthe-
sis, in: Topics in Current Chemistry, Vol. 206, (Ed.: P.
Knochel), Springer-Verlag, Berlin, 1999, pp 80–104;
b) A. Ogawa, D. P. Curran, J. Org. Chem. 1997, 62,
450–451.
[3] D. P. Curran, A. Solovyev, M. Makhlouf Brahmi, L.
Fensterbank, M. Malacria, E. Lacꢁte, Angew. Chem.
2011, 123, 10476–10500; Angew. Chem. Int. Ed. 2011,
50, 10294–10317.
[4] C. Chatgilialoglu, M. Newcomb, in: Advances in Or-
ganometallic Chemistry, Vol. 44, (Eds.: R. West, A. F.
Hill), Academic Press, San Diego, 1999, pp 67–112.
[5] X. Pan, E. Lacꢁte, J. Lalevꢀe, D. P. Curran, J. Am.
Chem. Soc. 2012, 134, 5669–5675.
[6] C. Chatgilialoglu, M. P. Bertrand, C. Ferreri, in: Sulfur-
Centered Radicals, (Ed.: Z. B. Alfassi), Wiley, West
Sussex, 1999, pp 311–354.
3526
asc.wiley-vch.de
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 3522 – 3526